TW201423883A - Method and apparatus for autonomous tool parameter impact identification system for semiconductor manufacturing - Google Patents

Abstract

A system and method for autonomously determining the impact of respective tool parameters on tool performance in a semiconductor manufacturing system is provided. A parameter impact identification system receives tool parameter and tool performance data for one or more process runs of the semiconductor fabrication system and generates a separate function for each tool parameter characterizing the behavior of a tool performance indicator in terms of a single one of the tool parameters. Each function is then scored according to how well the function predicts the actual behavior of the tool performance indicator, or based on a determined sensitivity of the tool performance indicator to changes in the single tool parameter. The tool parameters are then ranked based on these scores, and a reduced set of critical tool parameters is derived based on the ranking. The tool performance indicator can then be modeled based on this reduced set of tool parameters.

Description

Method and apparatus for automated tool parameter impact identification system for semiconductor manufacturing

The present invention is generally directed to techniques for determining the relative impact of tool parameters on a selected tool performance indicator of a semiconductor manufacturing system.

Technological advances in advances in electronics and computing devices have spurred advances in semiconductor technology. The ever-increasing consumer demand for smaller, more efficient, and more efficient computer devices and electronic devices has led to the need for semiconductor devices to scale down. In order to meet the device requirements while limiting the cost, the size of the germanium wafer on which the semiconductor device is formed has been increased.

Manufacturers of large wafer sizes use automation to implement and control wafer processing. These plants are capital intensive and there is therefore a need to maintain efficient operation of manufacturing equipment to minimize downtime and maximize throughput. To facilitate these goals, measurement equipment is typically employed during wafer processing to monitor manufacturing equipment and to obtain measurement information on the equipment and processed wafers. This measurement information is then analyzed to optimize the manufacturing equipment.

According to an example, the measurement information may include tool level information indicating the status or condition of the manufacturing device or a portion thereof, wafer metrology information specifying the physical and/or geometric conditions of the wafer to be processed, electronic document information, and the like. In addition, spectral data (eg, spectral line intensity information) can be collected to facilitate process engineers to identify the etch endpoint. However, in conventional manufacturing environments, various measurement data are processed independently of each other for different purposes. Therefore, the interrelationship between various measurement data cannot be balanced against further optimization of the manufacturing process.

The above-mentioned details of today's semiconductor manufacturing measurement and optimization systems are merely an introduction to some of the problems of conventional systems, rather than an exhaustive description. Other problems with the conventional system and the corresponding advantages of the various non-limiting embodiments described herein can be gained by reading the following description.

The following presents a simplified summary to provide a basic and general understanding of some aspects of the illustrative, non-limiting embodiments described herein. This summary is not an extensive overview of the invention, and is not intended to identify the main/critical elements or the scope of the various aspects described herein. Rather, the abstract is intended to be illustrative of the present invention as a

One or more embodiments of the present invention are directed to techniques for automatically identifying the relative impact of tool parameters on selected tool performance indicators in a semiconductor manufacturing system. To this end, a parameter impact identification system is provided that can balance the measured tool parameter data and tool performance data to identify The most important tool parameters that affect the efficiency of a particular tool. The parameter impact recognition system can also rank these important parameters based on their relative impact on the selected instrument's efficiency, providing useful guidance to maintenance personnel to identify which important tool parameters should be used to optimize the tool's performance metrics for the selection. The focus of maintenance work.

The parameter impact recognition system can determine the relative impact of the respective tool parameters by independently analyzing each tool parameter, and for each parameter, utilize the independent parameter to attempt to predict the behavior of a selected tool performance indicator. This can reduce the complex dimensions of semiconductor tool parameters to a single input single output (SISO) problem, where the tool performance indicator can be described as being a function of only a single tool parameter. The effect of each parameter on the performance indicator can then be determined based on an analysis of the resulting function (eg, by calculating a derivative of each function, by determining the prediction accuracy of each function, etc.), and the tool Parameters are sorted by relative impact.

In some embodiments, the parameter impact recognition system can also generate a function that characterizes the selected tool performance indicator in accordance with only the most important tool parameters determined by the above ranking. By eliminating tool parameters that have a slight impact on performance indicators, the resulting function greatly simplifies the problem space for end users and enables a sharper focus on these important tool parameters.

In order to achieve the above and related ends, some of the illustrative aspects described herein are combined with the following description and the accompanying drawings. These patterns are indicative of the various methods that can be implemented, all of which are covered herein. Other advantages and novelty features when reading with the schema You can get an understanding from the detailed instructions below.

100‧‧‧ system

102‧‧‧Input wafer

104‧‧‧Processed Wafer

106‧‧‧User Specifications

108‧‧‧Tool parameter data

110‧‧‧Semiconductor Manufacturing System

112‧‧‧Tool performance data

120‧‧‧ Spectrometer

130‧‧‧Tool Sensor

140‧‧‧Device measuring equipment

150‧‧‧Report component

154‧‧‧ Analysis results

160‧‧‧Parameter Impact Identification System

202‧‧‧Parameter Impact Identification System

204‧‧‧Interface components

206‧‧‧Parameter separation component

208‧‧‧Quality appraisal components

210‧‧‧ Sensitivity components

212‧‧‧Sort components

214‧‧‧Filter components

216‧‧‧synthetic function component

218‧‧‧ processor

220‧‧‧ memory

302‧‧‧Semiconductor Manufacturing System

304‧‧‧Tool parameter data

306‧‧‧Tool performance data

308‧‧‧Parameter Impact Identification System

310‧‧‧Parameter separation component

312‧‧ User specifications

314‧‧‧ Sensitivity component

316‧‧‧Filter components

318‧‧‧Sort components

320‧‧‧Composite function components

322‧‧‧Synthesis function

324‧‧‧Interface components

326‧‧‧Quality appraisal components

400‧‧‧ interface

500‧‧‧ interface

602‧‧‧Isolation parameter function

702‧‧‧Quality comparison

802‧‧‧ Sensitivity rating

902‧‧‧Parameter evaluation

904‧‧‧ sorted tool parameters

906‧‧‧Higher sorting tool parameters

908‧‧‧Lower sorted tool parameters

1002‧‧‧Top tool parameters

1100‧‧‧User interface

1102‧‧‧Information field

1104‧‧‧Information field

1402‧‧‧Data repository

1406‧‧‧System Memory

1420‧‧‧Application

1500‧‧‧ method

1502‧‧‧Steps

1504‧‧‧Steps

1506‧‧‧Steps

1508‧‧‧Steps

1510‧‧‧Steps

1600‧‧‧ method

1602‧‧‧Steps

1604‧‧‧Steps

1606‧‧‧Steps

1608‧‧‧Steps

1610‧‧‧Steps

1612‧‧‧Steps

1614‧‧‧Steps

1616‧‧‧Steps

1618‧‧‧Steps

1700‧‧‧ Computing environment

1702‧‧‧ Computer

1704‧‧‧Processing unit

1706‧‧‧System Memory

1708‧‧‧System Bus

1710‧‧‧ Non-volatile memory

1712‧‧‧ Random access memory

1714‧‧‧Disk storage

1716‧‧‧ interface

1718‧‧‧Operating system

1724‧‧‧Program Module

1726‧‧‧Program data

1728‧‧‧Wired/wireless input devices

1730‧‧‧Input device (interface)埠

1734‧‧‧Output (Adapter)埠

1736‧‧‧ peripheral output devices

1738‧‧‧Remote computer

1740‧‧‧Memory/storage device

1742‧‧‧Network interface

1744‧‧‧Adapter

1800‧‧‧ computing environment

1802‧‧‧User side

1804‧‧‧Server

1806‧‧‧Communication Architecture

1808‧‧‧Client Data Repository

1810‧‧‧Server Data Repository

Figure 1 is a block diagram showing an exemplary system for collecting and analyzing information about semiconductor manufacturing.

Figure 2 is a block diagram of an exemplary parameter impact identification system that can automatically identify tool parameters that affect the efficiency of the selected tool.

Figure 3 is a block diagram showing the processing functions performed by an exemplary parameter affecting recognition system.

Figure 4 illustrates an exemplary interface for selecting one of the tool performance indicators to be analyzed.

Figure 5 illustrates an exemplary interface for selecting one or more tool parameters to be considered by the parameter-affecting recognition system.

Figure 6 depicts a set of given tool parameters and performance data to produce a set of isolation parameter functions.

Figure 7 illustrates the assignment of quality ratings to the respective tool parameters based on the isolation parameter function.

Figure 8 illustrates the assignment of sensitivity ratings to the respective tool parameters based on the isolation parameter function.

Figure 9 illustrates the sorting of tool parameters in accordance with the relative impact on a tool performance indicator.

Figure 10 illustrates filtering the sorted tool parameters to identify one of the most important tool parameter sets that have the highest impact on a tool performance indicator.

Figure 11 illustrates an exemplary interface for configuring tool parameter filtering criteria.

Figure 12 graphically summarizes the identification of important tool parameters in accordance with one or more embodiments of the parameter impact identification system.

Figure 13 illustrates the generation of a composite function that characterizes a tool performance behavior in accordance with a reduced set of important tool parameters.

Figure 14 illustrates updating the tool performance function in a continuous manner.

Figure 15 is a flow diagram of an exemplary method for modeling a functional relationship between a tool performance indicator and a set of tool parameters of a semiconductor manufacturing system.

Figure 16 is a flow diagram of an exemplary method for automatically identifying and modeling the effects of tool parameters on a tool's efficiency.

Figure 17 is an example computing environment.

Figure 18 is an example network connection environment.

The invention will be described with reference to the drawings, in which the same elements are used to indicate the same elements. In the following description, numerous specific details are set forth However, it will be apparent that the invention may be practiced without these specific details. In other instances, well-known structures and devices are shown in the form of a block diagram to facilitate the description of the invention.

The term "object" as used in this specification and drawings, "Module", "Interface", "Component", "System", "Platform", "Engine", "Selector", "Manager", "Unit", "Storage", "Network", " "producer" or the like is used to refer to a computer-related entity having a particular function or an entity associated with or part of an operating machine or device; such entities may be a combination of hardware, hardware, and firmware, Firmware, combination of hardware and software, software or software in execution. Further, entities identified by the above terms are collectively referred to herein as "functional elements." For example, a component can be, but is not limited to, a program running on a processor, a processor, an object, an executable, a thread, a program, and/or a computer. For purposes of explanation, an application running on a server and the server can be used as a component. One or more components can reside in a program and/or thread, and a component can be fixed on a computer and/or distributed between two or more computers. Moreover, these components can be executed from a variety of computer readable storage media having various data structures stored thereon. The components can communicate via the local and/or remote program, such as in accordance with one of the one or more data packets (eg, from one component to another component in a local system, a decentralized system) Interactions and/or information that interacts with other systems via signals such as one of the Internet networks). By way of example, a component can be a device having a specific function provided by a mechanical portion operated by an electrical or electronic circuit, the circuit being operated by a software or firmware application executed by a processor, wherein The processor can be internal or external to the device and execute at least a portion of the software or firmware application. As another example, a component can be a device that provides a particular function via an electronic component without a mechanical portion, which can include a processor therein. A software or firmware that performs at least some of the functions of the electronic component is performed. The interface can include input/output (I/O) components and associated processor, application or API (application interface) components. Although the examples presented above are directed to a component, the illustrative features or aspects are also applicable to objects, modules, interfaces, systems, platforms, engines, selectors, managers, units, storage, networks, and the like.

Furthermore, the term "or" is used to mean an inclusive "or" rather than an exclusive "or". That is, "X employs A or B" is used to mean any natural inclusive permutation unless otherwise indicated or can be seen from the context. That is, if X employs A; X employs B; or X employs both A and B, then in the case of any of the above examples, "X employs A or B" is satisfied. In addition, the articles "a" and "an" are used to mean "one or more", unless otherwise specified or clear from the context. Learn about the singular type.

Moreover, the term "set" as used herein excludes an empty set; for example, a set that does not have any elements therein. Thus, one of the "collections" in the present invention includes one or more elements or entities. As an illustration, a component set includes one or more components; a variable set includes one or more variables; and so on.

Various aspects or features will be presented in a system that can include several devices, components, modules, and the like. It should be understood and appreciated that various systems may include additional devices, components, modules, etc., and/or may not include all of the devices, components, modules, etc. discussed in connection with the drawings. Wait. A combination of these methods can also be used.

1 is a block diagram showing an exemplary system 100 for collecting and analyzing information relating to semiconductor manufacturing. As shown in FIG. 1, a semiconductor fabrication system 110 can receive an input wafer 102 and an output processed wafer 104. In an exemplary, non-limiting embodiment, semiconductor fabrication system 110 can be an etch tool that is removed from input wafer 102 via an etch process (eg, wet etch, dry etch, plasma etch, etc.) The unmasked material is produced from a processed wafer 104 having recesses and features formed thereon. Semiconductor fabrication system 110 can also be a deposition tool (eg, atomic layer deposition, chemical vapor deposition, etc.) that deposits material on input wafer 102 to obtain processed wafer 104.

A variety of different measurement devices, such as spectrometer 120, tool sensor 130, and device measurement device 140, can monitor the programs executed by semiconductor fabrication system 110 to obtain completely different information about the various aspects, conditions, or results of the program. . By way of example, spectrometer 120 can acquire spectral intensity information including a set of intensities for each of the respective wavelengths or spectral lines observed by spectrometer 120. The spectral intensity information can be time series data such that the spectrometer 120 measures the intensity for each wavelength at regular intervals (eg, every 1 second, every 2 seconds, every 100 milliseconds, etc.). Spectrometer 120 can also correlate spectral intensity information to wafer IDs associated with a given wafer processed by semiconductor fabrication system 110. Thus, spectrometer 120 can individually acquire spectral intensity information for each wafer processed by semiconductor fabrication system 110.

Processing input wafer 102 at semiconductor fabrication system 110 While the corresponding sensor information is generated, the tool sensor 130 can monitor and measure tool operating characteristics. Tool sensor information, similar to the spectral intensity information measured by spectrometer 120, can be associated with each wafer based on time series data. Tool sensor information can include measurements from a variety of different sensors. Such measurements may include, but are not limited to, pressure in one or more chambers of semiconductor fabrication system 110, gas flow rate for one or more different gases, temperature, upper radio frequency (RF) power, self The time elapsed after the final wet cleaning, the aging of the tool parts, and the like.

Device measurement device 140 can measure the physical and geometric characteristics of the wafer and/or features that are fabricated on the wafer. For example, device measurement device 140 can measure development inspection critical dimension (DI-CD), final inspection critical dimension (FI-CD), etch bias, thickness, and the like at a predetermined location or region of the wafer. Measurement characteristics can be aggregated and output as device measurement information based on each location, each wafer. The characteristics of the wafer are usually measured before or after processing. Therefore, compared to spectral intensity information and tool sensor information, device measurement information is typically time series data acquired at different intervals.

By balancing the data collected from measurement devices 120, 130, and 140 during wafer production, a large number of tool performance indicators can be measured or derived for semiconductor fabrication system 110. Exemplary performance indicators may include overall production statistics such as wafer throughput, system downtime, system uptime, and the like. The performance indicator may also include a metrology output, or a measurement characteristic of the final semiconductor wafer produced by the system, including, but not limited to, edge deviation, deposition thickness, number of particles (eg, Degree of contamination), sidewall angle or other such measurement characteristics. Certain performance indicators may also consider data external to the semiconductor manufacturing system 110. For example, system-related repair costs may be calculated based in part on recorded financial or billing information obtained from one or more commercial-grade servers.

The behavior of a given tool performance indicator is typically a function of one or more tool parameters measured for the system. Tool parameters that have an impact on tool performance may include, for example, chamber pressure, temperature, RF power, gas flow, or other parameters measured during tool operation. Other tool parameters that affect tool performance include tool operating characteristics such as part aging, processing time, tool set-up time, wafer loading or unloading time, and more. Some tool performance indicators may also be one of a function of one or more product metrology inputs (eg, incoming critical dimensions, deposited thickness, refractive index of the material, or other such weights).

Since tool performance is primarily a function of one or more of the tool parameters described above, knowing which tool parameters have the greatest impact on the efficiency of a given tool will provide the user with a greater degree of control over the performance of the tool, and Helps maintain tool performance indicators within the required limits. However, since tool parameter data is usually only slightly associated with tool performance data, it is difficult to obtain which tool parameters have the highest impact on the efficiency of these tools. This information is useful for device owners to identify where maintenance should be focused.

To address these issues, a parameter impact identification system 160 is provided that can balance tool parameter data and tool performance data to automatically identify which tool parameters are associated with a selected tool performance indicator. Have the biggest impact. The parameter impact identification system 160 can also characterize the tool performance indicator for a small number of tool parameters that are determined to have a significant impact on the performance indicator, thereby simplifying the analysis by reducing dimensional complexity.

The parameter impact identification system 160 can receive the tool parameter data 108 and the tool performance data 112 as inputs. In one or more embodiments, the input data can be derived from a tool program record table that records parameters and performance data measured during the respective processes of semiconductor manufacturing system 110. The tool record table may include measurement data from one or more of spectrometer 120, tool sensor 130, or device measurement device 140. Measurement records in such tool program records may include, but are not limited to, sensor readings (eg, pressure, temperature, power, etc.), maintenance related readings (eg, aging of the focus ring, Aging of the mass flow controller, the time since the last maintenance was performed, the time since the last batch was loaded with the photoresist, etc.), and/or tool and performance statistics (eg, wafer processing time, chemicals) Consumption, gas consumption, etc.).

In an exemplary depiction, a tool program record table can be generated by a report component 150 at the end of each process flow of semiconductor fabrication system 110. At the end of a process flow, data from one or more of spectrometer 120, tool sensor 130, or device measurement device 140 may be provided to report component 150, which collects the collected data for the process A tool program records the table. A tool program record table may correspond to processing a single semiconductor wafer during the process, or corresponding to a batch of semiconductors fabricated during the process. The utility log can then be stored for use as a report or archive. From the tool The tool parameter data 108 and the tool performance data 112 of the sequence record table can be automatically provided to the parameter impact identification system 160 by an operator manually or by the reporting component 150 or a related device.

Although the above example describes the tool parameter data 108 and the tool performance data 112 retrieved or retrieved from the tool program record table, it should be understood that this data may also be provided to the parameter impact recognition system 160 by other means. For example, in some embodiments, all or a subset of tool parameter data 108 or tool performance data 112 may be provided directly from device 120, 130 or 140 to parameter impact identification system 160.

Tool parameter data 108 may include values (eg, pressure, temperature, power, gas flow, etc.) measured for one or more tools during operation, operational performance statistics (eg, part aging or usage times, processing time, installation) Time, load or unload time, etc., or other such tool parameters. Tool performance data 112 may include measurement characteristics of the final semiconductor wafer that are affected by one or more tool parameters (eg, etch deviation, deposition thickness, number of particles, sidewall angle, etc.), performance data for the tool itself (eg, crystal Other such metrics for round production volume, downtime, uptime, repair costs, etc., or operational performance of the tool.

The parameter impact identification system 160 processes the tool parameter data 108 and the tool performance data 112 in view of the user defined user specification 106. The user specification 106 can specify one or more processing preferences, including (but not limited to) the selection of tool performance indicators to be analyzed, the tool parameters to be considered (eg, which tool parameters are to be compared to the selected tool performance). The indicator is associated with) the preferred number of top tool parameters to be recognized by the system, preferably Learning methods (eg, simulated annealing, symbol regression, etc.) or other user preferences. The parameter impact identification system 160 analyzes the tool parameter data 108 and the tool performance data 112 in accordance with the user specification 106 to generate an analysis result 154, which will be described in greater detail below. In general, the analysis results 154 help identify important tool parameters that have the greatest impact on a given tool performance indicator, characterizing the selected tool performance indicator as a function of one of the important tool parameters, and The future value of the tool performance indicator giving the identified tool parameters is predicted.

Figure 2 is a block diagram of an exemplary parameter impact identification system that affects the identification system to automatically identify tool parameters that affect the performance behavior of the selected tool. The aspects of the system, apparatus, or program described in this disclosure may constitute a machine-executable component embodied in a machine, such as one or more computer-readable media (or media) associated with one or more machines. ) specific implementation. Such components, when executed by one or more machines (eg, computers, computing devices, automated devices, virtual machines, etc.), can cause the machine to perform the operations described above.

The parameter impact identification system 202 can include an interface component 204, a parameter separation component 206, a quality evaluation component 208, a sensitivity component 210, a sequencing component 212, a filtering component 214, a composite function component 216, and one or more processes. The device 218 and the memory 220. In many embodiments, interface component 204, parameter separation component 206, quality rating component 208, sensitivity component 210, sequencing component 212, filtering component 214, synthesis function component 216, one or more processors 218, and one of memory 220 Or more may be electrically and/or communicatively coupled to each other to perform the parameter influence One or more of the functions of system 202. In some embodiments, components 204, 206, 208, 210, 212, 214, and 216 can include software instructions stored on memory 220 and executed by processor 218. The parameter impact identification system 202 can also interact with other hardware and/or software components not shown in FIG. For example, processor 218 can interact with one or more external user interface devices, such as a keyboard, mouse, display monitor, touch screen, or other such interface device.

The interface component 204 can be configured to receive input from a user of the parameter impact identification system 202 and provide output to the user. For example, the interface component 204 can present an input display screen to the user via any suitable input mechanism (eg, a keyboard, touch screen, etc.) that displays the screen prompting the user to the specification and accepts the user from the user. And other specifications. The parameter separation component 206 can be configured to generate a function that isolates the effects of each individual tool parameter to determine the effect of each tool parameter on a selected tool performance indicator. Each function that attempts to predict the behavior of the selected tool performance indicator is a function of a single tool parameter. The quality rating component 208 can be configured to evaluate the accuracy of the actual behavior of the tool performance indicator in accordance with the function of the parameter to evaluate each tool parameter. Sensitivity component 210 can be configured to determine the sensitivity of the selected tool performance indicator to each tool parameter based in part on a function generated by parameter separation component 206. The ordering component 212 can be configured to order the tool parameters in accordance with the respective effects of the tool performance indicators to be analyzed by the tool parameters (eg, as determined by the quality rating component 208 or the sensitivity component 210). Filter component 214 can be configured to determine the tool performance One or more tool parameters with the least impact on the indicator are excluded from consideration. The composite function component 216 can be configured to generate a function that describes the tool performance indicator as a function of the reduced set of tool parameters after the filter component 214 has excluded the least impact parameter. The one or more processors 218 can perform one or more of the functions described herein with reference to the systems and/or methods disclosed. The memory 220 can be a computer readable storage medium that stores computer executable instructions and/or information to perform such functions as described herein with reference to the systems and/or methods disclosed.

Figure 3 is a block diagram showing the processing functions performed by the recognition system by an exemplary parameter. As described above in connection with FIG. 1, parameter impact identification system 308 receives tool parameter data 304 and tool performance data 306 associated with one or more processes of semiconductor manufacturing system 302 as inputs. Tool parameter data 304 and tool performance data 306 may be provided automatically to parameter impact identification system 308 (e.g., by report component 150 of FIG. 1) or manually provided by the user via interface component 324 to the system.

In addition to tool parameter data 304 and tool performance data 306, parameter impact identification system 308 also receives user specifications 312 from a user via interface component 324. As described above, the user specification 312 can specify which tool performance indicators are to be analyzed, which tool parameters are to be considered in terms of their respective effects on the selected tool performance indicator, and one of the selected number of tops to be identified by the system. Tool parameters (or one of the selected minimum number of influential tool parameters to be identified and excluded), a preferred learning method (eg, simulated annealing, symbol regression, etc.), or other User preferences.

An exemplary non-limiting interface for defining one or more user specifications will be described in conjunction with Figures 4 and 5. Exemplary interfaces 400 and 500 can be presented to a user via interface component 324. The illustrative interface 400 of FIG. 4 can be used to select a tool performance indicator to be analyzed by the system. The exemplary interface 500 of FIG. 5 can be used to select one or more tool parameters to be considered by the parameter impact identification system 308. That is, the parameter impact identification system 308 will evaluate the relative impact of the tool parameters selected via the illustrative interface 500 on the tool performance indicators selected via the illustrative interface 400. The exemplary interfaces 400 and 500 allow tool performance indicators and tool parameters to be selected using a checkbox; however, any suitable technique for inputting tool performance and tool parameter specifications can be employed and this is within the scope of the present invention. Within the scope of one or more embodiments.

Returning now to Figure 3, the processing operations performed on tool parameter data 304, tool performance data 306, and user specification 312 will be described. After tool parameter data 304, tool performance data 306, and user specification 312 have been provided to parameter impact identification system 308, parameter separation component 310 will each tool parameter (eg, each selected for analysis by user specification 312). The tool parameters are separated and repeatedly attempt to predict the behavior of the selected tool performance indicator using each of the separate individual tool parameters. This procedure can be explained in more detail with reference to Figure 6. In this example, the tool parameters P0-PN are considered, where N is an integer greater than zero. Thus, tool parameter data 304 includes measurements for each tool parameter P0-PN for one or more process flows of semiconductor manufacturing system 302. Parameter data. Tool performance data 306 includes corresponding measurement data for a selected one or more process flows for a selected tool performance indicator.

The parameter separation component 310 checks the tool parameter data 304 and the tool performance data 306 to generate a set of isolation parameter functions 602, each of which corresponds to a single one of the tool parameters P0-PN. Each of the isolation parameter functions 602 characterizes the predicted behavior of the tool performance indicator (denoted as output O) according to a single tool parameter. The resulting isolation parameter function corresponding to the tool parameter P0-PN can be expressed as follows: O=f ₀ (P0) (1)

O=f ₁ (P1) (2)

...

O=f _N-1 (PN-1) (3)

O=f _N (PN) (4)

These functions establish a non-linear functional relationship between the selected tool performance indicator (output O) and each tool parameter P0-PN. For the above equation, the tool performance indicator output O is the same for each function, but will be described by a different function f ₀ -f _N (each being a function of the tool parameters P0-PN). Thus, parameter separation component 310 reduces the complex dimensionality of semiconductor tool parameters to a single input single output (SISO) sub-problem.

The parameter separation component 310 can utilize any suitable learning method to learn the nonlinear functional relationship f for each parameter P0-PN, including (but not limited to) genetic stylization, symbol regression, simulated annealing, neural networks, or Other such nonlinear function recognition systems. In one or more embodiments, the user may select a preferred learning method for use by parameter separation component 310 to derive functions f ₀ (P0)...f _N (PN) . In such embodiments, the selection of a preferred method of learning may be accomplished by interface component 324. The parameter separation component 310 can also be configured to repeatedly update the functions f ₀ (P0)...f _N (PN) when the new tool parameter data 304 and/or the tool performance data 306 are received.

Returning now to Figure 3, after the parameter separation component 310 has established a functional relationship f ₀ (P0)...f _N (PN) describing a functional relationship between each tool parameter and the tool performance indicator, the resulting function may Submit for important parameter identification. To facilitate identification of important tool parameters that have the greatest impact on selected tool performance indicators, parameter impact identification system 308 can utilize one or more of quality rating component 326, sensitivity component 314, filtering component 316, and sequencing component 318.

The quality evaluation component 326 will be described in detail with reference to FIG. The quality rating component 326 can determine a quality rating for each of the isolation parameter functions f ₀ (P0)...f _N (PN) by comparing the predicted output O of each function to the actual tool performance data 306. The resulting set of quality ratings 702 indicates that the isolation function of each tool parameter matches the accuracy of the tool performance behavior predicted by the actual tool performance data 306. For example, when a new tool parameter profile 304 is received for a parameter P0-PN after a new processing flow of one of the semiconductor fabrication systems 302, the quality rating component 326 can run out of the new via the corresponding isolation function f0(P0) of the parameter. The parameter value P0 is used to determine the value (value O) of the tool performance indicator predicted by the isolation function of the parameter. The quality rating component 326 can then compare this predicted value O to the actual value indicated by the tool performance data 306 and assign a quality rating to the parameter P0 to indicate how close the predicted value O is to the actual value of the tool performance indicator. . The quality rating component 326 repeats this rating procedure for each of the remaining parameters P1-PN to derive a set of quality ratings 702. In one or more embodiments, the quality rating component 326 can update the quality rating 702 in a repeating manner as new processing flow data is received.

The quality rating 702 can provide a measure for determining the relative impact of each parameter P0-PN on the tool performance indicator being analyzed. In general, its isolation parameter function predicts that a tool performance output O closely matches a tool parameter of the actual tool performance indicator may have a higher degree of impact on the tool performance indicator. Conversely, its isolation function repeatedly and incomprehensibly predicts that one of the actual tool performance tool parameters is less likely to have a correlation with the tool parameter indicator. Therefore, quality rating 702 reflects these relative impact levels.

Other techniques can also be used to determine the relative impact of tool parameters on tool performance indicators. For example, some embodiments of the parameter impact identification system 308 may include a sensitivity component 314 or a replacement thereof in addition to the quality evaluation component 326. As shown in FIG. 8, sensitivity component 314 can evaluate each isolation parameter function 602 and assign a sensitivity rating 802 to the respective tool parameters P0-PN based on this evaluation. Similar to quality rating 702, sensitivity rating 802 indicates the relative impact of each tool parameter on the tool performance indicator being analyzed. In general, the sensitivity rating for a given parameter is measured by one of the sensitivity of the tool performance indicator to changes in the tool parameters.

In one or more embodiments, sensitivity component 314 can generate the sensitivity rating 802 for each isolated parameter function f ₀ (P0)...f _N (PN) based in part on a numerical or symbolic differential calculation. For example, for an isolation parameter function O=fi(Pi) corresponding to one of the tool parameters Pi, the sensitivity component 314 can utilize a value or symbol differentiation to calculate a differential:

This differential system represents the rate of change of the predicted value O of the tool performance indicator corresponding to the change in the tool parameter Pi, which is a measure of the sensitivity of the tool parameter indicator to changes in the tool parameter Pi. Sensitivity component 314 can generate sensitivity rating 802 based in part on respective differential calculations for each tool parameter P0-PN.

Since the parameters P0-PN will typically represent different types of tool parameters represented by different engineering units and having different operating ranges, there is a need for the sensitivity component 314 to normalize these differentials in some way to accurately compare the respective sensitivities. Sensitivity component 314 can also consider the effective operating range of each tool parameter when deriving a sensitivity rating based on differentiation. For example, if a given tool parameter is known for its upper and lower operating limits, sensitivity component 314 may only consider portions of the differential curve of the parameter between the two operational limits when calculating the sensitivity rating of the parameter. In general, any suitable technique or method for deriving a sensitivity estimate for a tool parameter based on the differentiation of a parameter's isolation parameter function is within the scope of one or more embodiments of the present invention.

One or more embodiments of the parameter impact identification system 308 Each parameter P0-PN can be evaluated using only one of the quality rating component 326 or the sensitivity component 314. Other embodiments may include both the quality rating component 326 and the sensitivity component, and generate a rating for each parameter based on the synthesis of the quality rating and the sensitivity rating. In the latter approach, parameter impact identification system 308 can combine the quality and sensitivity ratings using any suitable combination technique. For example, the parameter impact identification system 308 can apply a weighting factor (or vice versa) to the quality rating of a tool parameter based on the sensitivity rating of the parameter. In another example, the two ratings can be added together. These combined techniques are merely exemplary, and any suitable technique for generating a synthetic rating based on the quality and sensitivity rating is within the scope of one or more embodiments of the present invention.

Once a set of parameter ratings has been obtained, the sorting component 318 can sort the tool parameters based on the ratings. FIG. 9 illustrates an exemplary parameter ordering performed by the ranking component 318. The parameter rating 902 for each tool parameter P0-PN may include the quality rating, sensitivity rating, or one of the two ratings as described above. These ratings represent the relative impact or impact of each tool parameter on the tool performance indicator being analyzed. Based on these ratings 902, the ranking component 318 sorts the parameters P0-PN according to their effect on the tool performance indicator.

The resulting tool parameter ordering 904 can be used to identify a first sort of tool parameter subset 906 that has a significant impact on the tool performance indicator, and a second lower ordering tool that has little or no impact on the tool performance indicator. Parameter subset 908. Based on this determination, the parameter impact identification system 308 can exclude the tool performance indicator by The effect of the lower ordering tool parameter subset 908 can be ignored to reduce the dimensional complexity of the tool performance indicator analysis. Accordingly, filter component 316 can receive the sorted tool parameters 904 generated by sort component 318 and exclude the tool parameter subset 908 from consideration, as shown in FIG. The remaining set of top tool parameters 1002 represents those tool parameters that are identified as having a significant impact on the tool parameter indicator.

One or more embodiments of the parameter impact identification system 308 may allow the user to specify the number of tool parameters for the bottom sort to be excluded, or the number of tool parameters to be retained at the top. This provides the extent to which the user controls the dimensional complexity of subsequent analysis. FIG. 11 illustrates an exemplary user interface 1100 for making this selection. The exemplary user interface 1100 can include a selectable data entry field for the user to specify the number of top parameters to be retained (data field 1102), or the number of bottom parameters to exclude (data field 1104) ). Filter component 316 can utilize these selection configurations to filter the sorted tool parameters 904 and output a list of top tool parameters 1002 accordingly.

Figure 12 briefly illustrates the tool parameter identification process performed by the parameter impact recognition system 308. For each tool parameter P0-PN considered, an isolation parameter function is generated that defines a non-linear functional relationship between the tool parameters and an analyzed tool performance indicator. These functions can be expressed as f ₀ (P0)...f _N (PN) . Output O represents the predicted value of the selected tool performance indicator as a function of only a single tool parameter. For example, the output O corresponding to one of the functions of the parameter Pi represents the predicted value of one of the tool performance indicators as a function of only one of the tool parameters Pi.

The function f ₀ (P0)...f _N (PN) is then evaluated based on the degree of influence or influence of the corresponding tool parameter on the determination of the tool performance indicator. The rating may be generated based on the accuracy of the respective output O of the tool performance indicator matching the actual measured value, the value of the respective function f ₀ (P0)...f _N (PN) or the symbolic differential calculation or a combination of these techniques. Next, the parameters P0-PN are ordered based on these ratings. The top ordered parameter 906 can then be identified and retained for further analysis, while the bottom ordered parameter 908 (which represents a tool parameter that has normal or no effect on the tool performance parameter) is excluded from consideration.

Once the top-ordered tool parameters are identified, the parameter impact recognition system 308 can present the results to the user (eg, via the interface component 324). By providing the user with a list of tool parameters that are determined to have the greatest impact on the tool performance indicator, the parameter impact identification system 308 can provide guidance on where maintenance should be focused to maintain the tool performance indicator Within the required operational limits.

In one or more embodiments, parameter impact recognition system 308 can perform further analysis on the top-ordered tool parameters to provide additional insight into the relationship between tool parameters and tool performance indicators. In particular, once the number of tool parameters P0-PN has been reduced to a identified subset of important parameters, the output O can be re-learned by the system based on the change in the reduced set of tool parameters. To this end, the system can include a composite function component 320 configured to generate a new function that characterizes the tool performance indicator output (O _NEW ) to one of the most important tool parameters identified by the filter component 316. function. As shown in FIG. 13, the composition function component 320 receives the top tool parameter 1002 identified by the filter component 316 as having the greatest impact on the tool performance indicator, and learns a new synthesis function 322 that changes based on the top tool parameter 1002. To predict the tool affects the parameter output O _NEW . For example, for one of the top-ordered tool parameters set P1, P8, P0..., the composite function component 320 can generate a composite function 322 having one of the following general formulas: O _NEW = f(P1, P8, P0...) (6)

The composite function component 320 can employ any suitable non-linear function recognition technique to learn the composite function, including, but not limited to, genetic stylization, symbol regression, neural networks, least squares fit, or other suitable technique.

The compositing function 322 greatly simplifies the analysis of the tool performance indicator by reducing the problem space to a smaller set of important tool parameters to allow the user to focus more on these important parameters. The compositing function can be balanced in a number of ways to facilitate analysis of a selected tool performance aspect relative to the tool parameters to determine the behavior of the performance aspect. For example, the new tool parameter data can be analyzed based on the synthesis function 322 to predict future values of the tool performance indicator. If one or more tool parameters begin to deviate due to degradation, the expected future value of these tool parameters can be manipulated by synthesis function 322 to determine when the tool performance indicator _{ONEW is} expected to fall outside of acceptable performance limits. . In this manner, the synthesis function 322 acts as a basis for approaching the immediate early warning system, which can identify when preventative maintenance should be performed and which tool parameters should be the focus of maintenance work. Synthesis analysis function 322 may also be more generally to provide an important tool for the prediction tool effectiveness indicator parameter insight into the relationship between the O _NEW. Thus, the parameter impact identification system 308 can act as an efficient function modeling system that can reduce the search space used to perform functional relationship modeling for a semiconductor manufacturing system.

In some aspects, parameter impact identification system 308 can perform a one-time, on-demand calculation of synthesis function 322 and/or a collection of tool program records provided for one of the systems (eg, provided by a user) A collection of system tool flow data) to sort. However, the parameter impact identification system 308 can also be configured to operate in a continuous, repetitive manner on a substantially instantaneous basis as new tool data is collected. This repeated processing is shown in Figure 14. The parameter impact identification system 308 can be configured to receive the tool parameter data 304 and the tool performance data 306 directly from a semiconductor manufacturing system when new data is acquired (eg, at the end of each tool flow), and based on the new data The composite function 322 is repeatedly updated in accordance with the user specification 312. The composite function 322 can be retained in a data repository 1402 and serves as the basis for a continuously updated model for predicting future tool performance, identifying important tool parameters that affect tool performance, and the like. In addition to recalculating the composite function 322, the parameter impact identification system 308 can also re-estimate the ordering of important tool parameters upon receipt of new tool material. In this way, the parameter sensitivity to tool deviation, tool maintenance, and other wear and tear can be taken into account by the continuous repetitive nature of individual tool parameter learning.

The parameter impact recognition system described herein can automatically identify and correlate model tool parameters with tool performance indicators. Sex without regard to the complexity of the tool. This is achieved by reducing the larger set of potentially relevant tool parameters to a smaller set of important tool parameters and modeling the relationship between these important parameters and the tool performance indicators. This system can be applied to many types of semiconductor fabrication tools including, but not limited to, plasma etching tools, atomic layer deposition tools, and chemical vapor deposition tools. This system can also generalize multiple tool performance outputs (eg, throughput, downtime, uptime, repair costs, etch variations, deposition thickness, number of particles, sidewall angles, etc.).

15-16 illustrate various methods in accordance with one or more embodiments of the present application. Although one or more of the methods illustrated herein are shown and described as a series of acts for the purpose of simplifying the description, it should be understood and understood that the invention is not limited to the The embodiments shown and described herein may occur in different orders and/or concurrently with other acts. For example, those skilled in the art will understand and appreciate that a method may alternatively be represented as a series of associated states or events, such as in a state diagram. Furthermore, not all illustrated acts may be required to implement a method in accordance with the invention. Moreover, when a completely different entity implements a completely different portion of the method, the interaction map can represent a methodology or method in accordance with the present invention. Still further, two or more of the disclosed exemplary methods can be implemented in combination with each other to achieve one or more of the features or advantages described herein.

Figure 15 illustrates an exemplary method 1500 for modeling a functional relationship between a tool performance indicator and a set of tool parameters for one of the semiconductor fabrication systems. In the beginning, in 1502, choose a semiconductor manufacturing A tool performance indicator of the system is analyzed. The selected tool performance indicator can represent tool performance output such as wafer throughput, system downtime, system uptime, repair costs, and the like. The tool performance indicator can also be a special tool output product characteristic such as etch deviation, deposition thickness, number of particles, sidewall angle, and the like.

In 1504, a set of tool parameters to be associated with the tool performance indicator is selected. Exemplary tool parameters can include metrology measurements for one or more fabrication wafers, such as input CD, deposited thickness, refractive index of the material, and the like. Tool parameters may also include sensor readings taken during the process (eg, pressure, temperature, power, gas flow, etc.) and/or tool performance measurements (eg, part aging on the tool, processing time, loading) Set time, wafer loading and unloading time, etc.).

At 1506, the relative impact of each tool parameter on the selected tool performance indicator is determined. The effect of a given tool parameter is a measure of the sensitivity of the tool's performance indicator to the given tool parameter, or the extent to which the given tool parameter affects the value of the tool's performance indicator. In 1508, one of the tool parameter subsets determined to have the greatest impact on the selected tool performance indicator is identified based on the relative impact determined in step 1506. In 1510, a functional relationship between the tool performance indicator and the subset of tool parameters identified in step 1508 is modeled. This method for modeling a semiconductor tool performance indicator can be greatly simplified by reducing the number of tool parameters to a smaller set of important parameters identified as having the greatest impact on the selected performance indicator. analysis.

Figure 16 illustrates the effect of automatically identifying and modeling tool parameters on a tool's energy efficiency. First, in 1602, tool parameter data and tool performance data are received. The data may correspond to one or more processing flows of a semiconductor manufacturing system. The tool parameters can include metrology measurements for one or more fabrication wafers, such as input CD, deposited thickness, refractive index of the material, and the like. Tool parameters may also include sensor readings taken during the process (eg, pressure, temperature, power, gas flow, etc.) and/or tool performance measurements (eg, part aging on the tool, processing time, loading) Set time, wafer loading and unloading time, etc.). The tool performance indicator can include information related to tool performance output, such as throughput, downtime, uptime, repair costs, and the like. The tool performance data may also include tool output product characteristics such as etch deviation, deposition thickness, number of particles, sidewall angle, and the like.

At 1604, the tool parameter data is separated for the respective N tool parameters. In 1606, a counter i is set to one. In 1608, a function O = fi(Pi) that characterizes the relationship between the tool performance indicator O and the tool parameter Pi is generated for the i-th tool parameter. The function can be learned based on tool parameter data for Pi (separated in step 1604) and tool performance information about the tool performance indicator.

In 1610, a determination is made as to whether a function has been generated for all N parameters. If it is determined that a function has not been generated for all N parameters, then the method moves to step 1612 where counter i is incremented and step 1608 is repeated for the next tool parameter. Alternatively, if it is determined in step 1510 that a function has been generated for all N parameters, the method moves to step Step 1614.

In 1614, each function generated by steps 1608-1610 is evaluated based on the degree of matching of the predicted value O actual tool performance data and/or based on the sensitivity measured for the derived function. The sensitivity measurement describes the sensitivity of the tool performance indicator to changes in the tool parameters and can be determined based in part on numerical or symbolic differential calculations for each function.

In 1616, the N tool parameters are ordered according to the ratings determined in 1614, and the M highest ranked tool parameters are identified. These M highest ranked tool parameters represent a subset of all tool parameters that have been determined to have the greatest impact on the instrumental energy. In 1618, a new function is generated that models the tool performance indicator as a function of the M highest ranking parameters identified in step 1616. This new function can be used to predict future tool performance behavior based on tool parameter trends, to help schedule preventive maintenance and identify where maintenance should be focused, or other applications.

Various different aspects (eg, in combination with receiving one or more choices, determining the meaning of the one or more choices, distinguishing a selection from other actions, implementing a selection to satisfy a request, etc.) may employ a variety of different artificial intelligence based approaches. Implement its various aspects. For example, a program for determining whether a particular action is for a request to perform an action or a normal action (eg, an action that the user wants to perform manually) may be implemented via an automatic classifier system and program.

A classifier is a function that takes an input attribute vector x = (x1, x2, x3, x4, xn) maps to one of the trustworthiness of the input, that is, f(x) = confidence (category). Such classification may employ probabilistic and/or statistical based analysis (eg, analysis factor decomposition success and cost) to judge prognosis or to infer that the user wants to perform one of the actions automatically. In selected examples, for example, the attributes of a primary chamber and/or a reference chamber can be identified and the category meets the criteria for the primary chamber and/or reference chamber that must be employed.

A support vector machine (SVM) can employ an example of a classifier. The SVM operates by finding a hypersurface in the space of possible inputs that attempts to separate the trigger criteria from non-triggering events. Intuitively, this would result in a classification correction of test data that is close but not identical to the training material. Other direct or indirect model classification methods that provide different independent types can be used, including, for example, naïve Bayes, Bayesian networks, neural networks, fuzzy logic models, and probability classification. model. The classifications used in this paper may also include statistical regression, which is used to develop a priority model.

As can be readily appreciated from this specification, one or more aspects may be subjected to explicit training (eg, by genetic training materials) and by implicit training (eg, by observing user behavior, receiving external information). Classifier. For example, the SVM is configured through a learning or training phase in a classifier builder and feature selection module. Thus, the classifier can be used to automatically learn and perform a number of functions, including, but not limited to, determining which chambers to compare and which chambers to group together, according to a predetermined criterion when comparing chambers. The relationship between the chambers and so on. The standard may include (but not This is limited to requests, historical information, and so on.

Referring now to Figure 17, a block diagram of a computer is shown that is operable to perform the disclosed aspects. In order to provide additional context for its various aspects, FIG. 17 and the following discussion are a brief, general description of one suitable computing environment 1700 for providing embodiments in which various aspects may be implemented. Although the above description is in the general context of computer-executable instructions that can be executed on one or more computers, those skilled in the art will appreciate that various embodiments can be combined with other programming modules and/or hardware. Implemented in combination with one of the software.

In general, program modules include routines, programs, components, data structures, and the like that perform specific tasks or implement specific summary data types. Furthermore, those skilled in the art will appreciate that the disclosed aspects can be implemented with other computer system configurations, including single processor or multiprocessor computer systems, minicomputers, mainframe computers, and personal computers, Handheld computing devices, microprocessor-based or programmable consumer electronics, microcontrollers, embedded controllers, multi-core processors, etc., each of which is operatively coupled to one or more related Connected devices.

The aspects depicted in the various embodiments can also be implemented in a decentralized computing environment, some of which can be performed by remote processing devices that are coupled through a communications network. In a decentralized computing environment, the program module can be located in both the local and remote memory storage devices.

Computing devices typically include a variety of different media, which may include computer readable storage media and/or communication media, the two terms used herein being different from one another hereinafter. Computer readable storage media can Any available storage medium that is accessible by a computer and includes both volatile and non-volatile media, removable or non-removable media. By way of example, and not limitation, the computer-readable storage medium can be implemented in conjunction with any method or technique for information storage, such as computer readable instructions, program modules, structured material, or unstructured material. The computer readable storage medium may include, but is not limited to, RAM, ROM, EEPROM, DRAM, flash memory or other memory technology, CD-ROM, multi-purpose optical disc (DVD) or other optical disc storage , magnetic card, magnetic tape, disk storage or other magnetic storage device, or other tangible and/or non-transitory media that can be used to store the desired information. The computer readable storage medium can be accessed by one or more local or remote computing devices for various operations relative to information stored by the media, such as via an access request, query, or other data retrieval protocol.

Communication media typically embody computer readable instructions, data structures, program modules, or other structured or unstructured material in a data signal, such as a modulated data signal, such as a carrier or other transport mechanism, and Includes any messaging or delivery media. The term "modulated data signal" means a signal that has one or more of its set of characteristics or that changes in the manner in which information is encoded in one or more signals. By way of example and not limitation, communication media includes wired media, such as wired or fixed-line connections, and wireless media such as acoustic, microwave, RF, infrared, and other wireless methods (eg, IEEE 802.12X, IEEE 802.15.4) ).

Referring again to FIG. 17, the computing environment 1700 for implementing various aspects includes a computer 1702 including a processing unit 1704, a system memory 1706, and a system bus. 1708. The system bus 1708 couples system components (including the system memory 1706, but not limited thereto) to the processing unit 1704. The processing unit 1704 can be any commercially available processor. Dual microprocessors and other multiprocessor architectures can also be used as the processing unit 1704.

The system busbar 1708 can be any number of types of busbar structures that can be further interconnected to a memory busbar (with or without a memory controller) using various commercially available busbar architectures, peripherals. Bus and local bus. The system memory 1406 includes a read only memory (ROM) 1710 and a random access memory (RAM) 1712. A basic input/output system (BIOS) is stored in a non-volatile memory 1710, such as a ROM, EPROM, EEPROM, which contains basic routines to facilitate the transfer of information between elements in the computer 1702, such as During the boot. The RAM 1712 can also include a high speed RAM for caching data, such as static RAM.

The computer 1702 includes a disk storage 1714, which may include an internal hard disk drive (HDD) (eg, EIDE, SATA), and the internal hard disk drive may also be configured for external use in a suitable chassis (not In the illustration), a disk drive (FDD) (for example, for reading or writing a removable magnetic card), and an optical disk drive (for example, reading a CD-ROM disc, or for reading or writing) Into other high-capacity optical media such as DVD). The hard disk drive, the magnetic disk drive and the optical disk drive can be connected to the system bus 1708 by a hard disk drive interface, a disk drive interface and an optical disk drive, respectively. Interface 1716 for an external drive implementation includes at least one or both of a universal serial bus (USB) and an IEEE 1094 interface technology. Other external drive connection technologies are described in this article. The idea of various different embodiments.

The drive and its associated computer readable medium provide non-volatile storage of data, data structures, computer executable instructions, and the like. For computer 1702, the drive and media can accommodate the storage of any data in an appropriate digital format. Although the above description of computer readable media refers to HDDs, removable disks, and removable optical media (such as CDs or DVDs), those skilled in the art will appreciate that other types of media that can be read by a computer, Such as compressed disks, magnetic cards, flash memory cards, ports, etc., may also be used in the illustrated operating environment, and further, any such media may contain computer executable instructions for performing the disclosed aspects.

A plurality of program modules can be stored in the drive and RAM, including an operating system 1718, one or more applications 1420, other program modules 1724, and program data 1726. All or part of the operating system, applications, modules and/or materials may also be cached in the RAM. It will be appreciated that the various embodiments can be implemented in various commercially available operating systems or combinations of operating systems.

The user can enter commands and information into the computer 1702 via one or more wired/wireless input devices 1728, such as a keyboard and indicator device (such as a mouse). Other input devices (not shown) may also include a microphone, an IR remote control, a joystick, a game pad, a stylus, a touch screen, and the like. These and other input devices are typically coupled to processing unit 1704 via an input device (interface) 埠 1730 coupled to system bus 1708, but may also be connected by other interfaces, such as parallel 埠, IEEE 1094 series埠, game 埠, USB 埠, IR interface, etc.

A monitor or other type of display device can also be coupled to system bus 1708 via an output (adapter) 埠 1734, such as a video adapter. In addition to the monitor, a computer typically includes other peripheral output devices 1736, such as speakers, printers, and the like.

The computer 1702 can operate via a logical connection in a networked environment via wired and/or wireless communication to one or more remote computers, such as a remote computer 1738. The remote computer 1738 can be a workstation, a server computer, a router, a personal computer, a portable computer, a microprocessor based entertainment device, a peer device or other shared network node, and typically includes a computer 1702 Many or all of the elements are illustrated, however, for the sake of brevity, only one memory/storage device 1740 is shown.

The remote computer can have a network interface 1742 that can be logically coupled to the computer 1702. The logical connection includes wired/wireless connections to a local area network (LAN) and/or a larger network, such as a wide area network (WAN). These LAN and WAN connectivity environments are common in offices and companies, and promote enterprise-class computer networks, such as corporate intranets, all of which can be connected to a global communications network, such as the Internet.

When used in a LAN network connection environment, the computer 1702 is connected to the regional network via a wired and/or wireless communication network interface or adapter (communication connection) 1744. The adapter 1744 can facilitate wired or wireless communication to the LAN, and can also include a wireless access point disposed thereon for communicating with the wireless adapter.

When used in a WAN network connection environment, the computer 1702 can include a modem, or be connected to one of the communication servers on the WAN, or have other means to establish communication with the WAN, such as by the Internet. A data machine (which may be internal or external and a wired or wireless device) is coupled to system bus 1708 via a serial port interface. In a networked environment, the program modules described with respect to computer 1702, or portions thereof, may be stored in remote memory/storage device 1740. It should be understood that the network connections shown are for illustration and other means for establishing a communication link between computers.

The computer 1702 is operable to communicate with any wireless device or entity operatively disposed in a wireless communication manner, such as a printer, scanner, desktop and/or portable computer, portable data assistant, communication Satellite and any objects (such as kiosks, newsstands, etc.) and telephones of the device or location associated with the wirelessly detectable tag. This includes at least Wi-Fi and Bluetooth( ^TM ) wireless technology. Thus, the communication can be as a predefined structure of a conventional network, or only wireless random network communication between at least two devices.

Wi-Fi or wireless compatibility allows connections to the Internet without wires. Wi-Fi is a wireless technology similar to that used in mobile phones, which enables such devices (such as computers) to send and receive data indoors and outdoors; as long as it is anywhere within the base station. Wi-Fi networks use a radio technology called IEEE 802.11x (a, b, g, etc.) to provide a robust, reliable, fast wireless connection. Wi-Fi networks can be used to connect computers to each other, to the Internet, and to wired networks (which use IEEE 802.3 or Ethernet).

Wi-Fi networks operate in the unlicensed 2.4 and 5 GHz radio bands. IEEE 802.11 is generally applied to wireless LANs and utilizes frequency hopping spread spectrum (FHSS) or direct sequence spread spectrum (DSSS) in the 2.4 GHz band to provide a 1 or 2 Mbps transmission rate. IEEE 802.11a is an extension of IEEE 802.11 that is applied to wireless LANs and provides transmission rates of up to 54 Mbps in the 5 GHz band. IEEE 802.11a uses an orthogonal frequency division multiplexing (OFDM) coding scheme instead of FHSS or DSSS. IEEE 802.11b (also known as 802.11 High Rate DSSS or Wi-Fi) is an extension of 802.11 that is applied to wireless LANs and provides 11 Mbps transmission rates (with 5.5, 2, and 1 Mbps back-feed) in the 2.4 GHz band. IEEE 802.11g is applied to a wireless LAN and provides a transmission rate of 20+ Mbps in the 2.4 GHz band. Products can contain more than one frequency band (eg, dual frequency), so the network can provide real-world performance similar to the basic 10BaseT wired Ethernet used in many offices.

Referring now to Figure 18, there is shown an architectural block diagram of an illustrative computing environment 1800 for processing one of the disclosed architectures in accordance with another aspect. The computing environment 1800 includes one or more clients 1802. Client 1802 can be hardware and/or software (eg, threads, programs, computing devices). The client 1802 can include, for example, information about various different embodiments and/or related contextual information.

Computing environment 1800 also includes one or more servers 1804. Server 1804 can also be hardware and/or software (eg, threads, programs, computing devices). The server 1804 can include threads, for example, with respect to various different embodiments to perform the conversion. On the client side 1802 and the server One of the possible communications between 1804 is in the form of a data packet that is transmitted between two or more computer programs. The data packet may include, for example, a message and/or related contextual information. Computing environment 1800 includes a communication infrastructure 1806 (e.g., a global communication network, such as the Internet) that can be used to facilitate communication between client 1802 and server 1804.

Communication can also be facilitated via wired (including fiber optic) and/or wireless technologies. The client 1802 is operatively coupled to one or more client data repositories 1808 for storing local information (eg, messages and/or related context information) of the client 1802. Similarly, the server 1804 is operatively coupled to one or more server data repositories 1810 for storing local information of the server 1804.

In addition to the various embodiments described herein, it is understood that other similar embodiments may be employed or modified and added to perform the same or equivalent functions of the corresponding embodiments without departing from the And other embodiments. Still further, multiple processing wafers or multiple devices may share the performance of one or more of the functions described herein, and as such, may be stored through a plurality of devices. Therefore, the present invention should not be limited to any single embodiment, but should be construed in accordance with the scope, spirit and scope of the appended claims.

The term "exemplary" as used herein is used as an example, instance, or illustration. To avoid confusion, the subject matter disclosed herein is not limited to such examples. In addition, any aspect or design described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other aspects or designs, and does not. The skilled artisan has equivalent exemplary structures and techniques known to those of ordinary skill. Furthermore, to avoid confusion, the terms "including", "having", "containing" and the like are used to refer to the term "comprising" as an open open word. It does not exclude any additional or other components.

The above system has been described for the relationship between several components. It should be appreciated that such systems and components can include such components or specified sub-components - some of the specified components or sub-components, and / or additional components, and various arrangements and combinations described above. Sub-components may also be implemented as components communicatively coupled to other components rather than being included in the main component (hierarchical). In addition, it should be appreciated that one or more components can be combined into a single component that provides a collection function or divided into a plurality of separate sub-components, and any one or more intermediate layers, such as a management layer, can be provided in communication. Coupled to these subcomponents to provide integrated functionality. Any of the components described herein may also interact with one or more other components not specifically described herein but generally known to those skilled in the art.

In view of the above exemplary system, the method that can be implemented according to the above-mentioned subject matter can also be understood by referring to the flowchart of each figure. Although the methods are shown and described in a series of blocks for purposes of brevity, it should be understood and understood that the various embodiments are not limited to the order of the blocks, as some blocks may be different The order depicted and described herein and/or occurs concurrently with other blocks. Although non-sequential or divergent processes are illustrated via a flow chart, it should be understood that various other differences and process paths of the blocks may be implemented. The same or similar results are achieved by path or sequence. Furthermore, not all illustrated blocks are required to implement the methods described herein.

1500‧‧‧ method

1502‧‧‧Steps

1504‧‧‧Steps

1506‧‧‧Steps

1508‧‧‧Steps

1510‧‧‧Steps

Claims (20)

A system for identifying an influence of a tool parameter on a tool performance in a semiconductor manufacturing system, comprising: at least one non-transitory computer readable medium having a computer executable component stored therein; and at least one processor executing A computer executable component stored in the at least one non-transitory computer readable medium: a parameter separation component configured to generate a set of functions, the function set being separately from a set of tool parameters a single tool parameter to characterize one of a tool performance indicator behavior; a sorting component configured to order the tool parameter set in accordance with one of the tool performance indicators to obtain a parameter ordering, wherein The relative impact is determined based on the set of functions; and a filter component configured to identify a subset of tool parameters that have a significant impact on the tool performance indicator based on the parameter ordering. The system of claim 1, further comprising a quality rating component configured to determine a quality rating for respective functions of the set of functions, wherein the quality rating is determined by the respective function predictions A relative measure of the proximity of one of the tool performance indicators to the actual value. The system of claim 2, wherein the sorting component is configured to be sorted according to the quality rating of the respective functions. This tool parameter collection. The system of claim 1, further comprising a sensitivity component configured to determine a sensitivity rating for respective functions of the set of functions, wherein the sensitivity rating is corresponding to the tool performance indicator A measure of the sensitivity of the change in the single tool parameter of the respective functions. The system of claim 4, wherein the sequencing component is configured to order the set of tool parameters in accordance with the sensitivity ranking of the respective functions. The system of claim 4, wherein the sensitivity component is configured to determine the sensitivity rating for the respective functions based in part on a value or symbol differentiation determined for the respective functions. A system as claimed in claim 1, further comprising an interface component configured to receive an input specifying a plurality of highest ranking tool parameters to be retained as a subset of the tool parameters or to be excluded A plurality of lowest ordering tool parameters to obtain at least one of the subset of tool parameters. The system of claim 7, wherein the interface component is configured to present the subset of tool parameters on a display. The system of claim 1, wherein the parameter separation component is configured to generate the function based on tool parameter data and tool performance data measured for one or more processing flows of a semiconductor manufacturing system. set. The system of claim 1, further comprising a composite function component configured to generate a composite function that characterizes the tool performance indicator according to the subset of tool parameters. The system of claim 10, wherein the composite function component is configured to repeatedly update the composite function based on at least one of a new tool parameter data or a new tool performance data. A method for determining an effect of one or more semiconductor tool parameters on a tool performance indicator, comprising: executing, by using at least one processor, computer executable instructions embodied on at least one non-transitory computer readable medium to perform a plurality of operations, the method comprising: deriving a set of functions based on tool parameter data and tool performance data recorded for a semiconductor manufacturing system, wherein the function set is separately characterized by a tool performance indicator and a tool parameter set a non-linear relationship between one of the single tool parameters; sorting the tool parameter set according to the degree of influence of the tool parameter of the tool parameter set on one of the tool performance indicators, wherein the degree of influence is determined based on the function set And identifying, based on the ranking, a subset of tool parameters that are determined to have a significant impact on the tool performance indicator. The method of claim 12, wherein the operations comprise: matching the tool based on one of the respective functions of the set of functions One of the proximitys of one of the performance indicators is determined to assign a quality rating to the respective function, wherein the ranking includes ranking the tool parameter sets based on the quality ratings for the respective functions. The method of claim 12, the operations comprising: determining, based on the sensitivity of the tool performance indicator to a change in the single tool parameter corresponding to a respective function of the set of functions, A sensitivity rating is assigned to the respective function, wherein the ranking includes ranking the tool parameter sets based on sensitivity ratings for the respective functions. The method of claim 14, wherein assigning the sensitivity rating comprises: performing a value or symbol differentiation on the respective function; and determining the sensitivity rating for the respective function based on the value or symbol differentiation. The method of claim 12, wherein the operations comprise deriving a synthesis function that characterizes a relationship between the tool performance indicator and the tool parameter subset. The method of claim 16, wherein the operations include repeatedly updating the composite function in accordance with at least one of a new tool parameter data or a new tool performance data. A computer readable medium having computer executable instructions stored thereon responsive to execution of a system including a processor causing the system to perform a number of operations, the operations Included: generating a set of functions that respectively describe a correlation between a single tool parameter of one of the tool parameter sets associated with a semiconductor manufacturing system and a tool performance indicator of the semiconductor manufacturing system; in accordance with the tool parameter set Sorting the tool parameter set by the respective tool parameters having a relative influence on one of the tool performance indicators, wherein the relative influence is determined based on the function set; and the output based on the sort is determined to have an apparent performance indicator for the tool A subset of tool parameters affected. The computer readable medium of claim 18, wherein the sorting comprises sorting the set of tool parameters based on respective quality ratings assigned to the set of functions, the respective quality ratings indicating that respective functions of the set of functions predict the The proximity of one of the tool performance indicators to the actual value. The computer readable medium of claim 18, wherein the sorting comprises sorting the set of tool parameters based on respective sensitivity ratings assigned to the set of functions, the respective sensitivity ratings indicating that the tool performance indicator is for The sensitivity of the change in the single tool parameter associated with the set of functions, respectively.